Laudando-Associates-LLC/d-fine-small

D-FINE Small

This repository contains the D-FINE Small model, a real-time object detector designed for efficient and accurate object detection tasks.

Try it in the Browser

You can test this model using our interactive Gradio demo:

Model Overview

• Architecture: D-FINE Small

• Parameters: 10.3M

• Performance:

  • mAP@[0.50:0.95]: 0.816

  • mAP@[0.50]: 0.983

  • AR@[0.50:0.95]: 0.859

  • F1 Score: 0.951

• Framework: PyTorch / ONNX

• Training Hardware: 2× NVIDIA RTX A6000 GPUs

Download

Format	Link
ONNX
PyTorch

Usage

To utilize this model, ensure you have the shared D-FINE processor:

from transformers import AutoProcessor, AutoModel

# Load processor
processor = AutoProcessor.from_pretrained("Laudando-Associates-LLC/d-fine", trust_remote_code=True)

# Load model
model = AutoModel.from_pretrained("Laudando-Associates-LLC/d-fine-small", trust_remote_code=True)

# Process image
inputs = processor(image)

# Run inference
outputs = model(**inputs, conf_threshold=0.4)

Evaluation

This model was trained and evaluated on the L&A Pucks Dataset.

License

This model is licensed under the Apache License 2.0.

Citation

If you use D-FINE or its methods in your work, please cite the following BibTeX entries:

@misc{peng2024dfine,
      title={D-FINE: Redefine Regression Task in DETRs as Fine-grained Distribution Refinement},
      author={Yansong Peng and Hebei Li and Peixi Wu and Yueyi Zhang and Xiaoyan Sun and Feng Wu},
      year={2024},
      eprint={2410.13842},
      archivePrefix={arXiv},
      primaryClass={cs.CV}
}